Search

In Japan, medical liquid-crystal display (LCD) and general LCD monitors have color temperatures of 7500 and 6500 K, respectively. The differences in color temperature make it difficult for radiologists to judge whether the same color is being displayed on the monitor. Therefore, the radiologist may overlook lesions. We examined chromaticity on a color scale test pattern to determine the relationships between color temperature (6500-12,500 K) of the medical color LCD monitors, there are three types of fluorescent light and three types of illuminance LCD monitors. As the color temperature of the monitor increased, the variation in chromaticity for grayscale test patterns increased and those variations for the blue scale test patterns decreased in a dark room and at 600 lux. In addition, even if the color temperature of the monitor was changed, the variation in chromaticity showed no change under fluorescent lighting with light bulb color and daylight color. The results of this study will be useful for quality control and quality assurance of medical LCD monitors in terms of illuminance and color temperature of the monitor.

We have been developing a computer-aided detection (CAD) scheme for pneumoconiosis based on a rule-based plus artificial neural network (ANN) analysis of power spectra. In this study, we have developed three enhancement methods for the abnormal patterns to reduce false-positive and false-negative values. The image database consisted of 2 normal and 15 abnormal chest radiographs. The International Labour Organization standard chest radiographs with pneumoconiosis were categorized as subcategory, size, and shape of pneumoconiosis. Regions of interest (ROIs) with a matrix size of 32 × 32 were selected from normal and abnormal lungs. Three new enhanced methods were obtained by window function, top-hat transformation, and gray-level co-occurrence matrix analysis. We calculated the power spectrum (PS) of all ROIs by Fourier transform. For the classification between normal and abnormal ROIs, we applied a combined analysis using the ruled-based plus the ANN method. To evaluate the overall performance of this CAD scheme, we employed ROC analysis for distinguishing between normal and abnormal ROIs. On the chest radiographs of the highest categories (severe pneumoconiosis) and the lowest categories (early pneumoconiosis), this CAD scheme achieved area under the curve (AUC) values of 0.93 ± 0.02 and 0.72 ± 0.03. The combined rule-based plus ANN method with the three new enhanced methods obtained the highest classification performance for distinguishing between abnormal and normal ROIs. Our CAD system based on the three new enhanced methods would be useful in assisting radiologists in the classification of pneumoconiosis.

If patient information, such as identification number or patient name, has been entered incorrectly in a picture archiving and communication system (PACS) environment, the image may be stored in the wrong place. To prevent such cases of misfiling, we have developed an automated patient recognition system for chest CT images. The image database consisted of 100 cases with present and previous chest CT images. A volume of interest (VOI) measuring 40 × 40 pixels was selected from the left lung region, bronchus region, and right lung region. Next, the overall lung region and these three regions in a current chest CT image were used as a template for determining the residual value with the corresponding four regions in previous chest CT images. To ensure separation between the same and different patients, we applied a combined analysis that employed the ruled-based plus artificial neural network (ANN) method. The overall performance of the method developed was examined in terms of receiver operating characteristic (ROC) curves. The performance of the rule-based plus ANN method using a combination of the four regions was higher than obtained using a rule-based method using these four regions separately. The automated patient recognition system using the rule-based plus ANN method achieved an area under the curve (AUC) value of 0.987. This automated patient recognition method for chest CT images is promising for helping to retrieve misfiled patient images, especially in a PACS environment.

The purpose of this study was to evaluate the detection performance of simulated nodules in chest computed tomography (CT) images and nuclear medicine images with an ordinary liquid crystal display (LCD) and a medical LCD (grayscale standard display function: GSDF) and gamma 2.2. We collected 72 chest CT image slices obtained from an LSCT phantom with simulated signals composed of various sizes and CT values and 78 slices of monochrome and color nuclear medicine images obtained from a digital phantom with a simulated signal composed of various sizes and radiation levels. Six observers performed receiver operating characteristic (ROC) analysis using a continuous scale. The area under the ROC curve (AUC) was calculated for each monitor. The average AUC values for detection of chest CT images on a medical LCD (GSDF), medical LCD (gamma 2.2), and ordinary LCD were 0.71, 0.67, and 0.73, respectively. The average AUC values for detection of monochrome nuclear medicine images using a medical LCD (GSDF), medical LCD (gamma 2.2), and ordinary LCD were 0.81, 0.75, and 0.72, respectively. The average AUC values for detection of color nuclear medicine images on a medical LCD (GSDF), medical LCD (gamma 2.2), and ordinary LCD were 0.88, 0.86, and 0.90, respectively. Observer performance for detection of simulated nodules in chest CT images and nuclear medicine images was not significantly different between the three LCD monitors. We therefore conclude that an ordinary LCD monitor can be used to detect simulated nodules in chest CT images and nuclear medicine images.

Misregistration errors occur at the periphery of the hepatic region due to respiratory- and interval-related changes in hepatic shape. To reduce these misregistration errors, we developed a temporal and dynamic subtraction technique to enhance small hepatocellular carcinoma (HCC) by using a 3D nonlinear image-warping technique. The study population consisted of 21 patients with HCC. We registered the present and previous arterial-phase CT images or the present nonenhanced and arterial-phase CT images obtained in the same position by 3D global-matching plus 3D nonlinear image-warping. Temporal subtraction images were obtained by subtraction of the previous arterial-phase CT image from the warped present arterial-phase CT image. Dynamic subtraction images were obtained by subtraction of the present nonenhanced CT image from the warped present arterial-phase CT image. When we used this new technique, the number of good or excellent cases increased from 14.2% (3/21 cases) to 71.4% (15/21 cases) on temporal subtraction images. With this technique, subjective rating scores for image quality improved in 57.1% of cases (12/21 cases) on temporal subtraction images and 81.0% of cases (17/21 cases) on dynamic subtraction images. The results indicated that the new subtraction images were greatly improved by use of the 3D nonlinear image-warping technique.

Pneumoconiosis is diagnosed as categories 0-4 according to the Pneumoconiosis Law. Physicians have difficulty precisely categorizing many chest images. Therefore, we have developed a computerized method for automatically categorizing pneumoconiosis from chest radiographs. First, we extracted the rib edge regions from lung ROIs. Second, texture features were extracted using a dot enhancement filter, line enhancement filter, and grey level co-occurrence matrix. Third, the rib edge regions were removed from these processed images. Finally, we used a support vector machine for feature analysis. In a consistency test, 56 cases (69.7%) were classified correctly, and 45 cases (61.8%) were classified correctly in a validation test. These results show that the proposed features and removal of the rib edge are effective in classifying the profusion of opacities that indicate pneumoconiosis.

When interpreting enhanced computer tomography (CT) images of the upper abdomen, radiologists visually select a set of images of the same anatomical positions from two or more CT image series (i.e., non-enhanced and contrast-enhanced CT images at arterial and delayed phase) to depict and to characterize any abnormalities. The same process is also necessary to create subtraction images by computer. We have developed an automated image selection system using a template-matching technique that allows the recognition of image sets at the same anatomical position from two CT image series. Using the template-matching technique, we compared several anatomical structures in each CT image at the same anatomical position. As the position of the liver may shift according to respiratory movement, not only the shape of the liver but also the gallbladder and other prominent structures included in the CT images were compared to allow appropriate selection of a set of CT images. This novel technique was applied in 11 upper abdominal CT examinations. In CT images with a slice thickness of 7.0 or 7.5 mm, the percentage of image sets selected correctly by the automated procedure was 86.6+/-15.3% per case. In CT images with a slice thickness of 1.25 mm, the percentages of correct selection of image sets by the automated procedure were 79.4+/-12.4% (non-enhanced and arterial-phase CT images) and 86.4+/-10.1% (arterial- and delayed-phase CT images). This automated method is useful for assisting in interpreting CT images and in creating digital subtraction images.